TTransferIQ

Read-only demo: a snapshot of the model's real predictions taken 2026-09-27. The full stack (PostgreSQL, FastAPI, training pipeline) runs locally with Docker; see the GitHub repo.

Football market intelligence

Real player valuations from a Random Forest model trained on historical performance and transfer-market data, with every prediction explained by real SHAP contributions, not a black box.

Players tracked

537

Real predictions issued

537

Top current valuation

€55.83m

Top valued players

View full rankings →

Honest confidence

Every prediction carries a real, model-derived range and an explicit confidence label — never dressed up to look more certain than the underlying data supports.

Real explainability

Each valuation is broken down by real SHAP contributions, including tackles and interceptions for defenders — not just goals and assists.

Kept current

Squad and club data reflects the most recently loaded Premier League season, not a stale snapshot from years ago.